Jagadeeshwar Sayaboina

Engineered a Centralized Payments System supporting multiple payment modes and subscription models, scaled to 500 TPS during peak traffic. Developed the Credit Guru AI agent using LangChain, FastAPI & AWS Bedrock, resulted in a 75% improvement in query resolution & a 20% rise in overall product adoption. Delivered a digital credit card onboarding and application flow, enabled diverse state transitions and increased customer acquisition by 20% within 6 months. Facilitated PCI-DSS compliance for ZET Card by resolving 300+ critical vulnerabilities, leveraged Amazon Inspector, structured code reviews & systematic remediation practices. Guided and mentored interns and SDE1 Engineers, ensured smooth onboarding and timely delivery of high-impact features. Delivered performance optimization initiatives across microservices & event-driven systems, achieved p99 latency <100ms, reduced bottlenecks & drove a 30% uplift in system efficiency.

Built high-efficiency data synchronization workers using Python, Kafka, Airflow, and BigQuery, reliably processed 50M+ daily user events. Led the migration from Google App Engine to GKE, re-architected services and optimized compute resources, generated an annual savings of $200K. Collaborated with Product, Data, and Partner Teams to translate functional needs into scalable, back-end-first architectural solutions. Improved system performance by migrating a monolithic service to gRPC-based microservices, achieved 24% higher throughput and 15% reduced latency. Sustained 84% unit test coverage, enhanced overall system stability & release confidence.

Achieved <90ms p99 latency for checkout and payment services, ensured high responsiveness during peak electric scooter purchase traffic. Automated invoice and audit workflow via event-driven data pipelines, reduced processing times from 48 hours to near real-time. Led the development of the customer support platform, implemented robust and consistent role-based access control (RBAC). Delivered a home-delivered test-ride platform, integrated a rules engine that drove a 16% increase in test ride-to-sale conversions.

Enhanced face recognition and identity-matching pipelines by refining Siamese network models using TensorFlow, data pickling, and multi-processing, resulted in faster and more reliable fraud detection workflow. Developed automated Aadhaar Masking and Document Classification solutions leveraging OCR frameworks (Tesseract, NER, Pyzbar), strengthened regulatory compliance and streamlined loan processing operations.